Bỏ qua

Agentic Decision Tree

The Agentic Decision Tree is the hierarchical structure of if-then-else tactical conditions that an elite player runs implicitly during a point — processed by the Basal Ganglia rather than the Prefrontal Cortex. It is what separates a player with "scripted tactics" from one who executes genuine, adaptive agency in real time.

Unlike a fixed tactical script ("always hit to the backhand"), the Agentic Decision Tree is a dynamic hierarchy that updates to environmental anomalies at match speed — without conscious deliberation.

Structure

The Agentic Player sets a Primary Goal (e.g., "Take time away") before the point begins. From that goal, a hierarchy of conditional branches unfolds during play:

  • If the opponent's ball is short → move in and attack the open court
  • If the opponent recovers cross-court → redirect down the line
  • If the opponent is out of position → finish with depth, not angle
  • If an anomaly occurs (unexpected bounce, pace shift) → branch to the adaptive sub-tree

Each branch fires implicitly. The player does not consciously evaluate conditions — the basal ganglia run the decision logic and trigger the appropriate Motor Engram at the correct moment.

Why It Must Be Implicit

The 150ms execution window eliminates the possibility of conscious tactical decision-making during shot execution. By the time the PFC evaluates a conditional branch and issues a directive, the contact point has passed.

The decision tree must be embedded through Reinforcement Learning and Deliberate Practice — specifically through Constraints-Led Approach training where the player discovers tactical solutions independently rather than being told them. "Skill Emergence" produced this way is robust under sympathetic arousal; consciously memorized tactical rules are not.

Switching Without Performance Drop

Switching tactical plans mid-point requires neurological switching — the suppression of one motor preparation sequence and the activation of another. This is governed by mu/beta oscillatory power in the motor cortex. Players trained via Constraints-Led Approach develop robust switching capacity; players trained on scripted tactics often freeze when their primary plan is disrupted.

  • Basal Ganglia
  • Implicit Control
  • Motor Engram
  • Deliberate Practice
  • Reinforcement Learning
  • Deliberate Fluctuation
  • C-to-I Transition